The invention relates generally to product failure analysis. More specifically, the invention pertains to a failed component search technique where the original failed product containing a plurality of components must be at least partially destroyed in order to gain access to the plurality of components.
Known techniques for identifying a cause of a multi-component product failure involve component swapping. One such prior method begins by identifying two samples of the product being analyzed—one exhibiting the problem (a “worst” product) and one that does not exhibit the problem (a “best” product sample). The “worst” and the “best” product samples are then disassembled, and a first component is swapped between the two samples. The samples are reassembled and tested for the previously observed failure in the “worst” sample. If the “best” product sample now exhibits the previously observed problem, then the component at fault has been identified. If the “best” is still failure-free, then the samples are again disassembled and a different component is swapped. This process continues until the “best” sample exhibits the problem, to identify the component or components at fault.
This prior approach will not work, however, for products which are at least partially destroyed when disassembled. Such products typically employ non-mechanical fasteners such as welded seams or seals which are destroyed upon disassembly or dismantling.